Establishment and Control of Behavioral Bias in Drosophila Melanogaster

Abstract

Typically, variability observed across populations of animals is considered to be experimental noise. However, it is clear that individuals display considerable idiosyncrasies in behavior that are stable over both short and long time scales. Understanding how individuality arises and is maintained within the nervous system is a major challenge in neuroscience. I have found that biases in individuality in locomotor decision making are shaped during development, especially during pupation. Disruption of normal synaptic partner matching expands the magnitude of average locomotor biases in a population of flies. These biases are the product of the motor system of the fly. A key pre-motor circuit in the brain, the central complex, is likely to be a significant contributor to these biases. Using anatomical evidence from extensive morphological studies performed by other groups, we modeled a core component of this circuit and found that these deep brain neurons robustly form a ring attractor network under a variety of parameters, providing evidence for this circuit as a tracking system of heading in the fly. Finally, we have found that these locomotor biases are not as stable as previously thought, they can be modulated by ambient luminance, being sometimes dramatically different in the light and dark. This context dependence is mediated by vision and neurons within the central complex, namely the same ring attractor network. Furthermore, we have found evidence that output neurons of this network directly bias locomotor behavior due to asymmetric synaptic output into a downstream area of the central complex, the lateral accessory lobe, in a context dependent manner.